1、Technology：Temporal and Contextual Trend-aware Transformer Push Notification Recommender

We propose a novel push notification recommendation model designed explicitly for news articles. It addresses issues related to time sensitivity, data sparsity, rapid iteration of candidate items, and privacy protection. The model utilizes multiple critical learners, integrating time and contextual information extractors and a trend-awareness learner to effectively extract relevant and vital information. This enables the model to evaluate the inertia of click behavior, extract users' temporal and contextual preferences, and determine the impact of recent trends. Additionally, we have developed a fusion function and a gate control network to flexibly and comprehensively extract user click preferences.

2、Inventor：Prof.  Che Lin (Graduate Institute of Communication Engineering, National Taiwan University)

3、Enterprise Eligibility：

(1)Industry Category： digital content industry

(2)Required Technical Know-how：data mining, AI system design

(3)Required Machinery and Equipment：deep learning

(4)Required Research or Technical Staff：1 person (or more)

(5)Other Requirements：No specific requirements

4、How to apply：Please go to the Center of Industrial-Academic Cooperation of the Office of Research and Development to inquire about relevant information.

5、Announcement period：Applications will be accepted from the announcement date.

6、Contact window:

Intellectual Property & Licensing Manager：Jeremy Lai

TEL：+886-2-3366-9949

Email：jeremylai@ntu.edu.tw

7、Please refer to the attachment for the technical disclosure information of this case, and download the documents required for the application from the link